The embodiment of the RSA used in the experiments eliminated a huge voltage difference between battery and sensor contacts. Originally, the voltage difference Λ was:Λ=VBAT_MAX−0.65V=4.2V−0.65V=3.55Vwhere VBAT_MAX is the maximum battery voltage.
FIG. 4 shows a schematic of an RSA 400. RSA 400 includes a battery/detection contact 444, an enable contact 450, and switches Q1 and Q2. Switches Q1 and Q2 are both MOSFET switches. Switch Q1 is similar to the second switch 348 in FIGS. 3A-3C for selectively coupling the battery contact 444 to a battery for charging the battery during the charging mode. Switch Q2 is similar to the first switch 336 in FIGS. 3A-3C for selectively switching the RSA 400 between the measurement mode and the detection mode. During normal operation of RSA 400, the MOSFET Q1 is OFF. Therefore, the battery contact 444 is disconnected from the battery during normal operation. At the same time, the battery contact 444 is biased with a pull-up resistor R17 to a +0.65V buffer 451. Accordingly, the battery contact and sensor contact 414 (VIN) are at the same potential. Theoretically, at these conditions there should be no leakage current between these contacts.
Four baseline tests were performed with three RSA units to prove this concept. The tests are explained below and followed by the results.
Test Setup
Test 1
The experimental setup for Test 1 on an RSA 500 is shown in FIG. 5. RSA 500 includes a battery/detection contact 502, a sensor contact 504, a ground contact 506, and an enable contact 508. In Test 1, the voltage from the battery/detection contact 502 to ground is measured during the measurement mode of RSA 500. Test 1 shows how well the RSA battery is isolated from the RSA contact. In theory, the voltage should be close to the BIAS voltage (0.65V).
Test 2
The experimental setup for Test 2 is shown in FIG. 6. Test 2 measures the voltage difference from the battery/detection contact 502 to the sensor contact 504 during the measurement mode. If RSA 500 is contaminated, this voltage may potentially cause the leakage current from the battery/detection contact 502. The target voltage difference is about zero Volts.
Test 3
The experimental setup for Test 3 is shown in FIG. 7. Test 3 measures the voltage from battery/detection contact 502 to ground in charging mode. Test 3 demonstrates the quality and controllability of the internal MOSFET switch that will be on when ENABLE contact 508 is pulled down. In theory, the voltmeter will show a full battery voltage.
Test 4
The experimental setup for Test 4 is shown in FIG. 8. Test 4 measures the leakage current between the battery/detection contact 502 and the sensor contact 504 with and without contamination. Test 4 shows the maximum theoretical leakage in case of contamination from the battery/detection contact 502. The ideal number is no more than a few pA.
Test Results
The results of Tests 1-4 for three different RSAs are shown in Table 1 below.
TABLE 1RSA #TESTTest DescriptionRSA 1RSA 2RSA 3TEST 1Battery to Ground0.6384 V0.6377 V0.6350 VVoltageTEST 2Battery to Sensor  0.0 mV  0.0 mV  0.0 mVVoltageTEST 3Battery to Ground 4.154 V 4.101 V 4.120 VVoltage in Charge 4.154 V 4.101 V 4.120 VmodeTEST 4Battery to Sensor  1-10 pA  1-5 pA  1-10 pALeakageAs shown in Table 1, all three RSAs have virtually no leakage between battery and sensor contacts.
However, despite the good test result, the voltage between battery contact 502 and sensor contact 504 is not always zero. Rather, the voltage between the battery contact 502 and sensor contact 504 is equal to the offset voltage of the operational amplifier in RSA 500. According to the datasheet of the operational amplifier used, its maximum offset voltage is ±150 μN. In comparison to the previous RSA approach, this voltage difference is small (±150 μV vs. 3.55V for old RSA). However, even this voltage difference may create some additional leakage that can affect the sensor current measurement.
FIGS. 9 and 10 show an RSA 900 with various currents illustrated to show how contamination may affect the measurement. RSA 900 includes a battery contact 902, sensor contact 904, ground contact 906, and enable contact 908. Switch Q1 (similar to the second switch 348 in FIGS. 3A-3C) is controlled by the enable contact 908 and selectively couples the battery contact 902 to a battery 910 for charging during a charging mode. Switch Q2 (similar to the first switch 336 in FIGS. 3A-3C) is controlled by a microprocessor 911 and selectively couples the battery contact to ground during a detection mode. In a measurement mode, both switches Q1 and Q2 are off, and the battery contact is at a bias voltage 912 (e.g., +0.65V).
FIG. 9 illustrates the RSA 900 in the measurement mode with both switches Q1 and Q2 off. A transimpedance amplifier 914 is biased with the bias voltage 912 to keep the voltage at the sensor contact 904 substantially the same as the battery contact 902 (+0.65V). Accordingly, a total sensor current 916 depends only on the resistance of sensor 918, with a leakage current 920 caused by contamination 921. According to the 1st Kirchhoff's Law, the total sensor current 916 is the sum of an amplifier input current 922 (e.g., bias current 922 in FIG. 9) and the leakage current 920. Accordingly, the input current 922 of the transimpedance amplifier 914 is affected by the leakage current 920 and the measurement may be incorrect.
FIG. 10 illustrates the RSA 900 in the detection mode with Q1 off and Q2 on. In a detection algorithm, the sensor measurement may be performed in two stages. A first measurement is performed when Q2 is OFF (as shown in FIG. 9). The second measurement is performed when Q2 is ON (as shown in FIG. 10). In the second stage, the voltage at the battery contact 902 drops down to about +4 mV. The voltage at the sensor contact remains +0.65V. Without a leakage current 920, both measurements will be substantially the same.
However, if a surface of the RSA 900 is contaminated, the leakage current 920 from the sensor contact (+0.65V) will flow to the battery contact (+4 mV) during the detection mode, as shown in FIG. 10. The total input current 922 of the transimpedance amplifier 914 will rise, causing the output voltage of the amplifier 914 to also rise. By comparing two sequential results, the existence or absence of contamination (and thereby leakage current 920) is detected.
Simulation of Contamination Detection in TINA-TI
The contamination detection procedure discussed above was simulated in TINA-TI. The schematic used for the simulation is shown in FIG. 11. The results from the simulation are shown in Table 2.
TABLE 2SimulationOA output (mV)SensorSensorContamination@ +150 uV@ −150 uV% OF CHANGEResistanceCurrentResistancedifferencedifference@ +150 uV@ −150 uV(MΩ)(nA)(MΩ)Q2 = OFFQ2 = ONQ2 = OFFQ2 = ONdifferencedifference5.421200.11860330018403300 77% 79%1.01850330018503300 78% 78%5.01850330018503140 78% 70%10.01850250018502500 35% 35%100.01850191018501910 3% 3%10.83600.11260330012403300162%166%1.01250330012503300164%164%5.01250254012503300103%164%10.01250190012501900 52% 52%100.01250131012501310 5% 5%1006.51.0724.543300709.273300355%365%5.0715.482010714.712100181%194%10.0715.251360714.851360 90% 90%100.0715.02779.6714.99779.6 9% 9%20,0000.3250.1659.873300644.63300400%412%(20 GΩ)(325 pA)1.0652.483300649.033300406%408%5.0650.811940650.031940198%198%10.0650.571300650.181300100%100%100.0650.31714.93650.31714.93 10% 10%
Additionally, FIGS. 12A-D show graphs that demonstrate how the transimpedance amplifier output voltage swings when MOSFET Q2 switches ON/OFF (thereby switching between detection mode and measurement mode, respectively) at different Sensor Resistance and contamination levels. As it was expected, the voltage change is more significant at a lower contamination resistance. The worst detection case is the highest sensor current and the lowest leakage. For instance, at 120 nA sensor current (5.42M—lowest sensor resistance) and 100M Contamination Resistance, the voltage changes by only about 3%. However, that change may be enough to detect the leakage current. The permissible leakage current may be determined based on a number of factors. The simulations shown in FIGS. 12A-D were done for the contamination resistance range of 100 kΩ-20 GΩ. For the values below 100 kΩ, the detection is even easier because the voltage may swing by more than 77% as shown.
Contamination Detection Algorithms
Based on the simulations, the detection algorithms shown in FIGS. 13 and 14 are proposed. FIGS. 13 and 14 refer to components of the RSA 900 as shown in FIGS. 9 and 10.
FIG. 13 shows a detection algorithm 1300. At block 1302, Q2 is turned off, thereby placing the RSA in the measurement mode. If Q2 is already turned off, block 1302 may be omitted. A sensor measurement is then performed at block 1304 to obtain a first value of the sensor signal received by the RSA during the measurement mode. At block 1306, Q2 is turned on, thereby placing the RSA in the detection mode. A sensor measurement is then performed at block 1308 to obtain a second value of the sensor signal received by the RSA during the detection mode.
At block 1310, the first and second values (Data1 and Data2, respectively) are transmitted to a monitoring unit. The monitoring unit receives the first and second values at block 1312. At block 1314, the monitoring unit compares the first value with the second value according to the formula (Data2−Data1)/Data1. The monitoring unit compares this quantity to a threshold (THOLD).
If the calculated quantity is less than the threshold, the monitoring unit uses Data1 for processing at block 1316 (e.g., for determining the glucose level of the patient). However, if the calculated quantity is greater than the threshold, the monitoring unit determines that contamination is present on the RSA contacts at block 1318. The monitoring unit initiates one or more actions in response to the finding of contamination, such as activating an alarm and/or excluding Data1 from being used for processing. In an aspect, data sets Data1 and/or Data2 may include additional values in addition the first value and second value, respectively. The additional data may be used for detection purposes, and/or for sensor measurement purposes.
In an aspect, the value of the threshold (e.g., the actual maximum allowable difference between the first and second values) may be determined based on tests with real contamination substances.
FIG. 14 illustrates a method 1400 in which the RSA performs the comparison between the first value and the second value, rather than the monitoring unit. Blocks 1402, 1404, 1406, and 1408 are similar to blocks 1302, 1304, 1306, and 1308, respectively, as discussed above. At block 1410, the RSA compares the first value to the second value and compares this quantity against a threshold. If the results of the comparison are less than the threshold, the RSA transmits the data to the monitoring unit at block 1412. In an aspect, the RSA only transmits Data1. Alternatively, the RSA may transmit Data2 and/or information related to the results of the detection algorithm 1400 in addition to Data1.
If the results of the comparison are greater than the threshold, the RSA determines, at block 1414, that contamination is present that is causing a leakage current. In response, the RSA generates an error code. The RSA transmits, at block 1412, data to the monitoring unit. The transmitted data may include Data1, Data2, the error code, and/or other information regarding the algorithm 1400. In an aspect, Data1 may not be transmitted to the monitoring unit if the results of the comparison are greater than the threshold.
The monitoring unit may include any suitable structure for communicating with the RSA, processing the data received from the RSA, and/or presenting information to the patient and/or a caregiver. For example, the monitoring unit may be a computing device, such as a personal data assistant, mobile phone, personal computer, laptop computer, tablet computer, and/or a dedicated computing device for the sensor system.
Although certain embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a wide variety of alternate and/or equivalent embodiments or implementations calculated to achieve the same purposes may be substituted for the embodiments shown and described without departing from the scope. Those with skill in the art will readily appreciate that embodiments may be implemented in a very wide variety of ways. This application is intended to cover any adaptations or variations of the embodiments discussed herein. Therefore, it is manifestly intended that embodiments be limited only by the claims and the equivalents thereof.